Regression Detection

このページは日本語には対応しておりません。随時翻訳に取り組んでいます。翻訳に関してご質問やご意見ございましたら、お気軽にご連絡ください。

Overview

A regression refers to the unintended reappearance of a bug or issue that was previously fixed. In Datadog, resolved issues are automatically re-opened through regression detection so that you can see all the context of the issues without duplicating information.

Automatic re-opening through regression detection

If a RESOLVED error recurs in a newer version of the code, or the error occurs again in code without versions, Error Tracking triggers a regression. The issue moves to the FOR REVIEW state, and is tagged with a Regression tag:

The details of regression in Error Tracking

Regressions take into account the versions of your service where the error is known to occur, and only trigger on new versions after an issue is marked as RESOLVED. Configure your services with version tags (see instructions for APM, RUM, and Logs) to ensure that issues automatically resolve only if the same errors occur on newer versions of your services.

If you don’t have version tags set up, issues are tagged with Regression when an error occurs on an issue marked as RESOLVED.

You can also set up monitors to alert you when regressions occur.

Updating the issue status

The issue status appears anywhere the issue can be viewed, such as in the issues list or on the details panel for a given issue. To manually update the status of an issue, click the status and choose a different one in the dropdown menu.

The Activity Timeline in the Error Tracking Issue

Issue history

View a history of your issue activity with the Activity Timeline. On the details panel of any Error Tracking issue, view the Activity Timeline by clicking the Activity tab.

The Activity Timeline in the Error Tracking Issue

Further Reading

お役に立つドキュメント、リンクや記事: